A NOBLE Trial and an Attempt to EXCEL: A Comparison of Two Randomized Trials of Unprotected LMT PCI vs. CABG

Editor's Note: This is Part One article of a two-part Expert Analysis. Click here for Part Two.

A quick look at the two recently published, well-conducted trials seems to show exactly the opposite conclusions. But do they?

NOBLE (Nordic-Baltic-British Left Main Revascularization Study)1 studied 1,200 patients randomized to biolimus-eluting stent versus coronary artery bypass graft surgery (CABG) and showed that CABG was superior to percutaneous coronary intervention (PCI) even for left main trunk (LMT) lesions of low SYNTAX scores. The primary endpoint of major adverse cardiac and cerebrovascular events (MACCE) was 29% for PCI versus 19% for CABG, even though the all-cause mortality was similar.

The EXCEL trial (Evaluation of XIENCE versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization)2 compared everolimus-eluting stent with CABG. The primary endpoint of MACCE was 15.4% for PCI and 14.7% for CABG. The conclusion is that the strategy of PCI for LMT, including bifurcation lesions of syntax scores <32, was non-inferior to CABG.

Both studies were randomized to PCI versus CABG for LMT lesions with syntax scores <32 in the majority of patients (including bifurcation lesions in 80% of patients). The patient characteristics were comparable, although the overall demographics were of Caucasian men aged 66 years, with incidence of diabetes ranging 15-30%. All patients had preserved left ventricular function.

Both studies had excellent procedural results, with 30-day mortality of about 1%. A total of 2.4 stents with a length of 48 mm was used in the EXCEL trial. The comparative data from NOBLE could not be discerned, but the mean stent diameter was 4.9 mm. In both groups, the majority had on-pump CABG with at least 93% use of left internal mammary artery to left anterior descending artery. In both studies, dual antiplatelet therapy was used in the PCI arm for at least 1 year, with aspirin in all CABG patients. A significant number of CABG patients also received clopidogrel if prior non-ST-segment elevation myocardial infarction (MI). The use of statins was in the mid 80% range.

NOBLE had a median 3.1-year follow-up after the initial pre-specified median 2-year data were not sufficient to elicit conclusions. The total follow-up reported was up to 5 years, with all patients followed for a minimum of 1 year. However, in the EXCEL trial, when the enrollment was not at target at 2 years to enroll the pre-specified 2,600 patients, the power calculations were modified with a median follow-up of 3 years (reported follow-up) to enroll a total of 1,900 patients. Data on follow-up beyond 3 years have not been reported for the EXCEL trial.

The endpoints were subtly different. NOBLE used a primary endpoint (MACCE) of all-cause mortality, cardiac and cerebrovascular events, non-procedural MI, and need for repeat revascularization during follow-up. Kaplan-Meier estimates of MACCE by intention-to- treat after 5 years were 29% (121 events) for PCI and 19% (81 events) for CABG. The hazard ratio was 1.48 (95% confidence interval, 1.11-1.96), exceeding the limit for non-inferiority (1.35), and was significant for superiority of CABG compared with PCI (p = 0.0066). The total non-procedural MI was 7% for PCI versus 2% for CABG at 5 years (hazard ratio 2.88; 95% confidence interval, 1.4-5.9; p = 0.004). The need for total revascularization was 16% for PCI versus 10% for CABG at 5 years (hazard ratio 1.5; 95% confidence interval, 1.04-2.17; p = 0.032).

The EXCEL trial had a primary endpoint (MACCE) of all-cause mortality, MI, and stroke but did not include need for myocardial revascularization at median of 3 years. The primary composite endpoint event of death, stroke, or MI at 3 years occurred in 15.4% of patients in the PCI group and 14.7% of patients in the CABG group (difference of 0.7 percentage points; upper 97.5% confidence limit, 4.0 percentage points, p = 0.02 for noninferiority; hazard ratio 1.00; 95% confidence interval, 0.79-1.26; p = 0.98 for superiority.)

In a post-hoc analysis of the period between 30 days and 3 years after randomization to PCI or CABG, more primary endpoint events occurred in the PCI group than in the CABG group. Ischemia-driven revascularization during follow-up was more frequent after PCI than after CABG (in 12.6 vs. 7.5% of the patients; p < 0.001). With longer follow-up, the need for ischemia-driven repeat revascularization seemed to increase in both studies for PCI arm more than in the CABG group.

Let's explore this issue of repeat revascularization and the impact on longer term mortality in the context of recently reported follow up data from the SYNTAX (Synergy Between Percutaneous Coronary Intervention With TAXUS and Cardiac Surgery) trial.3 The use of repeat revascularization as an endpoint can be controversial in clinical trials, potentially due to the subjective nature of these procedures. However, the need for repeat revascularization was an independent predictor of mortality, MI, and stroke after PCI compared with CABG for patients with LMT disease and 3 vessel coronary artery disease at 5 years.

In both studies, the length of hospitalization, need for transfusions, and risk of infection were higher for the CABG group compared with the PCI group. For both studies, however, the mortality rates for CABG at 30 days was remarkably low at 1%. Continued focus on reducing risk of infection (particularly deep sternal), reducing need for transfusions, optimally using arterial conduits (though the impact of this will likely not be evident until follow-up is more than 10 years), and applying best practices for medical management after CABG is essential.4,5

In summary, these well-done studies offer reassurance of the safety of PCI with modern generation stents in short-term follow-up. But before these data can be used to amend the current guidelines recommending CABG for managing LMT lesions, longer follow-up is essential.


  1. Mäkikallio T, Holm NR, Lindsay M, et al. Percutaneous coronary angioplasty versus coronary artery bypass grafting in treatment of unprotected left main stenosis (NOBLE): a prospective, randomised, open-label, non-inferiority trial. Lancet 2016;388:2743-52.
  2. Stone GW, Sabik JF, Serruys PW, et al. Everolimus-Eluting Stents or Bypass Surgery for Left Main Coronary Artery Disease. N Eng J Med 2016;375:2223-35.
  3. Parasca CA, Head SJ, Milojevic M, et al. Incidence, Characteristics, Predictors, and Outcomes of Repeat Revascularization After Percutaneous Coronary Intervention and Coronary Artery Bypass Grafting: The SYNTAX Trial at 5 Years. JACC Cardiovasc Interv 2016;9:2493-507.
  4. Taggart DP, Altman DG, Gray AM, et al. Randomized Trial of Bilateral versus Single Internal-Thoracic-Artery Grafts. N Engl J Med 2016;375:2540-9.
  5. Lamy A, Devereaux PJ, Prabhakaran D, Taggart DP et al. Five-Year Outcomes after Off-Pump or On-Pump Coronary-Artery Bypass Grafting. N Engl J Med 2016;375:2359-68.

Keywords: Coronary Artery Bypass, Coronary Artery Disease, Diabetes Mellitus, Drug-Eluting Stents, Hydroxymethylglutaryl-CoA Reductase Inhibitors, Myocardial Infarction, Myocardial Revascularization, Percutaneous Coronary Intervention, Stroke, Ticlopidine, Ventricular Function, Left, Angiography

< Back to Listings